A. ERICKSON, Koch Modular Process Systems, Paramus, New Jersey (U.S.); and B. CALDWELL, Koch Modular Process Systems, St. Louis, Missouri (U.S.)
Extractive distillation is a powerful technique for separating azeotropic and close-boiling mixtures in industrial processes. Acetonitrile and toluene, two common organic solvents, form an azeotrope that conventional distillation cannot break. Pressure swing distillation is also ineffective due to minimal composition shifts with pressure, making extractive distillation the preferred approach. Selecting the right solvent is crucial to modify relative volatilities and ensure efficient separation, but this choice impacts process feasibility, energy consumption and solvent recovery. This article presents technical considerations for selecting a solvent capable of efficiently and economically separating the acetonitrile/toluene azeotrope.
Extractive distillation. Conventional distillation methods rely on boiling point differences and cannot produce pure products from binary mixtures containing azeotropes. The acetonitrile/toluene azeotrope exemplifies this challenge, as its vapor and liquid compositions are identical at about 81 wt% acetonitrile at atmospheric pressure.1 This renders simple distillation ineffective. Pressure swing distillation, another conventional method, is impractical due to minimal composition shifts with pressure changes, leading to large capital expenditure and high energy consumption. Consequently, extractive distillation becomes the preferred approach for such separations. This article explores the principles of extractive distillation, the criteria for solvent selection and the impact on process performance, culminating in a case study on acetonitrile/toluene separation.
Limitations of conventional separation techniques. Distillation is the workhorse of chemical processing, widely used to separate mixtures based on differences in boiling points. While effective for many applications, it becomes ineffective when dealing with azeotropic mixtures. These systems exhibit constant boiling behavior where the vapor and liquid compositions are identical at a given pressure.2 Such mixtures cannot be separated by conventional distillation alone. Distillation also struggles with mixtures whose components have similar volatilities, when their relative volatility is close to one, meaning both components vaporize at nearly the same rate. In these cases, achieving separation requires many theoretical stages and high reflux ratios, resulting in substantial energy consumption and equipment size.
To address these limitations, alternative techniques such as pressure swing distillation, azeotropic distillation, extractive distillation or hybrid methods involving membranes, adsorption and pervaporation may be employed. However, these alternatives may have tradeoffs, including high capital and operating costs, limited effectiveness across pressure ranges, and challenges when handling thermally sensitive compounds.
Principles of extractive distillation. Extractive distillation is a separation technique used to handle azeotropic or close-boiling mixtures by introducing a high-boiling solvent. This solvent selectively alters the relative volatilities of the original components, disrupting azeotropic behavior or enhancing the separation of closely boiling species. By increasing the difference in volatility, the process allows one of the components of the azeotropic mixture to be distilled to higher purity than would be possible through conventional distillation. The solvent typically has a significantly higher boiling point than the feed components, facilitating its recovery and reuse.
Unlike azeotropic distillation, which relies on forming a new azeotrope with an added agent, extractive distillation avoids forming additional azeotropes. In azeotropic distillation, the solvent usually forms an immiscible phase and minimum boiling azeotrope, enabling separation by phase splitting. The phase separation requires a decanter, as well as accurate liquid-liquid interaction parameters for reliable process modeling.2 Operationally, these systems are more difficult to control compared to an extractive distillation system since two separate columns and condensers feed a single decanter, and two liquid phases are often present inside the column.
Pressure swing distillation separates azeotropic mixtures by exploiting the fact that azeotrope compositions can shift with pressure. When the azeotropic point changes significantly between two pressures, separation becomes feasible by operating two columns at different pressures.3 While this approach is conceptually elegant and avoids using a solvent, it depends on the azeotrope shifting meaningfully with pressure, which many do not. If the change in composition is too small, a large recycle between the two columns is required to achieve separation. In fact, this recycle rate approaches infinity as the two azeotropic compositions approach each other. This leads to significantly higher capital and energy costs.
Like pressure swing and azeotropic distillation, extractive distillation typically involves at least two columns, depending on the complexity of the solvent recovery process. As shown in FIG. 1, the first column receives two separate feeds: the "fresh feed," containing the solution to be separated, is introduced at the lower part of the column, while the solvent feed is added higher up. These feeds combine and are processed through three distinct sections of the extractive distillation column: solvent recovery, rectifying and stripping.
The solvent recovery section consists of trays between the solvent feed tray and the top of the column. Here, the component with higher relative volatility (the light product) is concentrated and exits as a distillate product. Reflux is necessary to separate this component from the solvent and provide reflux to the rectifying section. The rectifying section between the solvent feed and fresh mixture feed trays removes the component of lower relative volatility (the heavy product) with the solvent from the lighter component. The liquid and vapor composition at the top of this section is relatively free of the heavier product component, so the solvent always enhances the separation in a solvent-rich area. The stripping section between the mixture feed tray and the bottom of the column concentrates the heavier component (stripping off the lighter component), which is then removed as the bottom product.
Subsequently, the heavy component and the solvent are introduced into a recovery column. In this column, the solvent is separated as the bottom product and recycled back into the first column, while the distillate consists of a nearly pure heavy product.
Solvent selection criteria. Selecting an appropriate solvent is critical to the success of extractive distillation. The solvent must selectively alter the partial pressures and thus the relative volatilities of the mixture components without forming azeotropes or interfering with separation. The solvent should have a higher boiling point than the feed components to ensure it remains in the liquid phase during distillation and is easily recoverable. Additionally, the solvent should be chemically stable, non-reactive with feed components, non-toxic and highly selective in its interactions to favor the desired separation. Common solvents include glycols, glycerol, dimethyl sulfoxide and dimethylformamide, though the optimal choice depends on the specific feed composition and separation target.
An ideal solvent would increase the activity coefficient of the lower-boiling feed material and decrease the activity coefficient of the higher-boiling feed material. However, achieving both effects simultaneously is difficult. If a solvent accomplishes even one of these goals, or causes the system to behave ideally, it is considered a strong candidate. Ideally, the solvent should also be considerably heavier than the higher-boiling feed component to reduce the capital and energy requirements of its recovery. However, the solvent should not be so heavy that standard utilities such as 150-psig steam and cooling water cannot be used effectively for heating and condensation. Preliminary screening often involves generating an X-Y diagram for the binary system, with a fixed solvent concentration on a solvent-free basis to visualize separation feasibility. It is also essential to verify that the solvent does not form azeotropes with either of the feed components, as this could undermine the separation process. If the X-Y diagram shows improved separation behavior, the candidate solvent can be evaluated further through process simulation and pilot-scale studies. Solvent options are typically ranked based on their ability to maximize relative volatility and minimize the total annual cost of the process. FIG. 2 illustrates that adding phenol as a solvent to an isooctane/toluene mixture can yield favorable outcomes in extractive distillation. The X-Y plot shows the initial binary equilibrium (green curve) and the equilibrium in the presence of 70 mol% phenol solvent (blue curve).
CASE STUDY: ACETONITRILE/TOLUENE SEPARATION
The acetonitrile/toluene azeotrope presents significant separation challenges due to its azeotropic nature. To address this, the authors comprehensively screened potential solvents to identify a suitable candidate for extractive distillation. The authors identified a promising candidate after evaluating several solvents based on their ability to modify relative volatilities. Vapor-liquid equilibrium (VLE) tests were conducted with the selected solvent using proper procedures at 50 wt% and 70 wt% to prove its effectiveness as a solvent.4 Based on these results, a simulation model was developed to further verify that the solvent is an economically viable option for separating an acetonitrile/toluene mixture at > 90% purity and recovery.
FIG. 3 illustrates that adding the solvent to an acetonitrile/toluene mixture will yield favorable outcomes in extractive distillation. The X-Y plot shows the initial binary equilibrium (pink curve) and the equilibrium in the presence of 50 wt% and 70 wt% solvent (green and blue curves).
Takeaways. Extractive distillation is a robust and flexible solution for separating azeotropic and close-boiling mixtures that cannot be effectively processed using conventional distillation or pressure swing methods. In the case of acetonitrile and toluene, the lack of a significant composition shift with pressure rules out pressure swing distillation, making extractive distillation the most practical and economical choice. Success depends heavily on selecting the right solvent, which must modify relative volatilities without forming new azeotropes, remain chemically stable and operate within typical utility constraints.
This article demonstrates how solvent selection can dramatically impact process feasibility and economics through careful solvent screening, thermodynamic modeling and VLE testing. The presented case study highlighted a candidate solvent that enables separation of acetonitrile and toluene to high purity and recovery. HP
REFERENCES
Horsley, L. H., “Azeotropic data—III,” Advances in Chemistry, 1973, online: https://www.scribd.com/doc/309954120/Lee-H-Horsley-Azeotropic-Data-III
Gerbaud, V., I. Rodriguez-Donis, L. Hegely, P. Lang, F. Denes and X. You, “Review of extractive distillation: Process design, operation, optimization and control,” Chemical Engineering Research and Design, January 2019.
Wang, Y., Z. Zhang, Y. Zhao, S. Liang and G. Bu, “Control of extractive distillation and partially heat-integrated pressure-swing distillation for separating azeotropic mixture of ethanol and tetrahydrofuran,” Industrial & Engineering Chemistry Research, August 2015.
Schlowsky, G., A. Erickson and T. Schafer, “Generating your own VLE data,” Chemical Engineering, February 1995.
Alan H. Erickson is the Vice President for Koch Modular Process Systems. He is a chemical engineer with 45 yrs of experience in chemical manufacturing plants' process design, startup and project management. He has expertise in computer simulations and the design of unit operations such as evaporation, distillation, liquid-liquid extraction, absorption and adsorption, heat transfer, fluid flow, complete process control systems and instrumentation. He earned a BS degree in chemical engineering from Rutgers University.
BriCE Caldwell is a Process Engineer II at Koch Modular Process Systems. He has 8 yrs of experience in process engineering, with a background spanning the design and operation of chemical manufacturing facilities. He earned a BS degree in chemical engineering from Missouri University of Science and Technology, and an MS degree in mechanical engineering from the Georgia Institute of Technology.